Ah, I didn't see that you got it to work. It seems pretty strange that you
have to explicitly include this library in the config.h... Did you try
adding /usr/lib64 to LD_LIBRARY_PATH? I always thought the standard c++
library was linked by default... but maybe it's a CUDA compiler issue.
It's impossible for me to do more than speculate without being able to
reproduce that issue, but thanks for the report.
All the best,
Jason
On Wed, Aug 24, 2011 at 8:46 AM, Dmitry Mukha <dvmukha.gmail.com> wrote:
> the sequence of my actions was
> patch -p0 -N < bugfix.all.at15
> ./apply_bugfix.x.a11 bugfix.1to17.tar.bz2
> compiling of serial and parallel, with 'make clean' command after each step
> python AT15_Amber11.py
>
> and then ld linking will fail without additional flag(-s).
>
> 2011/8/24 Jason Swails <jason.swails.gmail.com>
>
> > Make sure all bug fixes are applied. If this is the recent patch
> > bugfix.1to17.tar.bz2, then you also need to fully patch AmberTools 1.5
> (the
> > Amber 11 patch should print out a message to that extent). There were
> > changes to configure necessary to the new pmemd.cuda patch.
> >
> > HTH,
> > Jason
> > >
> > > The solution was (applying after the configuration with -cuda flag)
> > > uname -a | grep 'fc14.x86_64' >/dev/null 2>&1
> > > if [ $? -eq 0 ]; then
> > > echo
> > > echo DM PATCH2 on a11
> > > echo -n "Patching $AMBERHOME/src/config.h "
> > > sed -i 's[PMEMD_CU_LIBS=-L$(CUDA_HOME)/lib64
> > > -L$(CUDA_HOME)/lib[PMEMD_CU_LIBS=-L$(CUDA_HOME)/lib64
> -L$(CUDA_HOME)/lib
> > > -L/usr/lib64 -lstdc++[' $AMBERHOME/src/config.h
> > > echo
> > > fi
> > >
> > > --
> > > Sincerely,
> > > Dmitry Mukha
> > > Institute of Bioorganic Chemistry, NAS, Minsk, Belarus
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > > http://lists.ambermd.org/mailman/listinfo/amber
> > >
> >
> >
> >
> > --
> > Jason M. Swails
> > Quantum Theory Project,
> > University of Florida
> > Ph.D. Candidate
> > 352-392-4032
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> Sincerely,
> Dmitry Mukha
> Institute of Bioorganic Chemistry, NAS, Minsk, Belarus
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Aug 24 2011 - 08:00:03 PDT
